From Prices to Parties: Downs

Downs notes in his introduction that, though government has an increasingly great impact on private-decision making, "little progress has been made toward a generalized yet realistic behavior rule for a rational government similar to the rules traditionally used for rational consumers and producers" (3).

Downs sets about remedying this by applying economic theory originally designed to explain price dynamics to Western Democratic party systems. Using a theoretic apparatus known as the "spatial market" developed by economist Harold Hotelling in the late 1920's, Downs seeks to explain several key features of party systems.

The Model and Its Assumptions

Downs begins by reducing all political questions to their bearing on one key issue: how much government intervention in the economy should there be? He then asks us to imaging a horizontal scale ranging from 1 to 100 representing the range of opinion on the subject. 1, the leftmost point on the scale represents full government control and 100, the right end, represents a completely free market. He concedes that this is unrealistic in that each party is leftish on some issues and extreme rightists tend to favor fascist control rather than free markets.

Hotelling, when he applied the model to politics, assumed that there were an equal number of voters at each point of the scale. Downs makes an important modification by adding a variable distribution of voters along the scale (a roughly normal distribution of voters, with the bulk of voters clumped in the middle, or for a more ideologically polarized city, a sharply bimodal distribution with peaks toward the two ends of the scale). We are also to assume, quite reasonably, that the primary objective of a party is to attract as many members as possible in an effort to win elections. Thus the platforms of parties, the number of parties, and the very stability of democracy are determined by the distribution of voters. In short, party ideology goes where the voters are. As Downs crisply puts it "the numerical distribution of voters along the political scale determines to a great extent what kind of democracy will develop" (121).

A few examples:

Voters' preferences are normally distributed (as in the U.S. case)

Add diagram here

If two parties begin ideologically at 25 and 27, they will tend to move toward the center, their policies becoming more and more similar, as they attempt to a portion of the large number of the voters in the middle. Also, as there are so few voters at either ideological extreme, the emergence of viable third parties will be rare.

Bimodally distributed

Where voters are bimodally distributed, the system will tend to be two-party, and these parties will remain ideologically distinct. Downs surmises that such system will be extremely unstable in that if one party comes to power it will attempt to implement policies that will be anathema to a significant proportion of the population. No government in this situation can operate so as to please most of the people and this situation may lead to revolution.

Evenly distributed

This sort of distribution can sustain a multiplicity of parties. But this arrangement also hinges on the existence of a proportional representation system. In a winner take all plurality system like that of the U.S. this would be unworkable because groups would have to coalesce in order to gain a sufficient number of votes to win a majority. But in a proportional system a far smaller number of votes is required to in order ensure a significant voice in government. Thus, parties in a multiple system will strive to distinguish themselves ideologically from each other and maintain the purity of their positions. The multiparty system can be thought of as occupying a middle ground between the two configurations described above. It is marked by significant ideological differences, but is quite stable.

More on electoral structure: it can be seen not only as interacting with the distribution of voters' preferences but also as a cause or an effect thereof.

Effect: the framers of the electoral structure may voter's preferences into account. If the distribution is unimodal, a plurality system may be chosen as a way of ensuring that this large group will not be politically ignored. And if the distribution has made small modes, law makers may opt for proportional representation to ensure that sizeable extremist groups will have a voice in government (Israel, would be a good example).

Cause: On the other hand, in a plurality structure, since it tends to produce just two, ideologically similar parties, voters' tastes may homogenize. A proportional structure, and the variety of ideologically distinct parties, may produce the opposite effect.

Third Parties

Downs identifies two types: those that form merely to try and influence the ideology of existent parties and those that actually seek to win elections. The first will form in an effort to pull a party away from the center toward one end of the political spectrum or the other. Downs challenges the view of these individuals as irrational, in that, as in the case of abstainers, are behaving in such a way that has little short-term return. Downs encourages us to think of abstainers and the fomenters behind such "blackmail parties" as exhibiting a future oriented rationality, there purpose being to alter the choices offered to voters by the extant parties at some future date.

Parties designed to actually win elections can only successful emerge when there occurs a marked shift in the distribution of voters' preferences. He uses the example of the emergence of Britain's labor party. It emerged with the extension of suffrage to the working class. Before 1990 the distribution of the electorate was unimodal toward the right, with two parties, as Downs' model predicts, Tory and Liberal. But upon the extension of the franchise created another mode to the extreme left, which allowed for the emergence of the Labor party and the ultimate demise of the more centrist Liberals

Overlap and Ambiguity

As noted above that in a two party system there occurs a great deal of overlapping of policy. As downs puts it: in the middle of the scale were most voters are massed, each party scatters its policies on both sides of the midpoint. It attempts to make each voters in this area feel that it is centered right at his position" (135). Now of course each party will throw in a few extreme stands in order to please its "far-out" voters. Therefore, in a two party system, it is usually only possible to tell which side of the middle line a party stands by examining its extremist policies.

In addition to this overlapping of policies in two party systems, the policies themselves are characterized by a great measure of ambiguity. Such policy ambiguity allows for a wide range of interpretations and thus allows each policy stand to cover a greater number of voters. It is in a party's interest to be as equivocal about their policies as possible. As a result, voters wind up making their choice on the basis of something other than the issues, such as the personality of candidates, traditional family voting patterns, etc. As Downs concludes, "rational behavior by political parties tends to discourage rational behavior by voters" because they can never know enough about a parties policies to make the choice that will best serve their interests.